Show simple item record

AuthorMelhem, Rawad
AuthorJafar, Assef
AuthorAl Dakkak, Oumayma
Available date2023-12-24T09:56:14Z
Publication Date2024
Publication NameThe Sixth Youth Research Forum 2024
URIhttp://hdl.handle.net/10576/50552
AbstractSpeech separation is very important in real-world applications such as human-machine interaction, hearing aids devices, and automatic meeting transcription. In recent years, a significant improvement occurred towards the solution based on deep learning. In fact, much attention has been drawn to supervised learning methods using synthetic mixtures datasets despite their being not representative of real-world mixtures. The difficulty in building a realistic dataset led researchers to use unsupervised learning methods, because of their ability to handle realistic mixtures directly. The results of unsupervised learning methods are still unconvincing. In this paper, a method is introduced to create a realistic dataset with ground truth sources for speech separation. The main challenge in designing a realistic dataset is the unavailability of ground truths for speakers’ signals. To address this, we propose a method for simultaneously recording two speakers and obtaining the ground truth for each. We present a methodology for benchmarking our realistic dataset using a deep learning model based on Bidirectional Gated Recurrent Units (BGRU) and clustering algorithm. The experiments show that our proposed dataset improved SI-SDR (Scale Invariant Signal to Distortion Ratio) by 1.65 dB and PESQ (Perceptual Evaluation of Speech Quality) by approximately 0.5. We also evaluated the effectiveness of our method at different distances between the microphone and the speakers, and found that it improved the stability of the learned model.
Languagear
PublisherQatar University Young Scientists Center - Qatar University
SubjectSingle Channel
Speech Separation
Deep Learning
Realistic Datasets
Ground Truths
TitleTowards Solving Cocktail-Party: The First Method to Build a Realistic Dataset with Ground Truths for Speech Separation
TypeConference Paper


Files in this item

Icon

This item appears in the following Collection(s)

Show simple item record